The proposal is concerned with the mechanism underlying two important phenomena characteristic of mitochondrial genetics in yeast (Saccharomyces cerevisiae). These phenomena, polarity and suppressiveness, are characterized by the net loss of mitochondrial gene markers contributed by one parent in zygotes heterozygous for the mitochondrial omega locus or for petite mutations. We will test a new, specific hypothesis of the mechanism for this loss. Our hypothesis states that such cells are heterozygous for deletions or duplications of mitochondrial genes; during recombination, heteroduplexes are formed in the mitochondrial DNA regions containing the deletions or duplications and a single-stranded loop is formed. The preferential excision and repair of this loop leads to the loss of the gene markers contained in the longer genome and the conversion of closely-linked markers. Experimental approaches are: (1) genetic experiments designed to test the postulate that recombination is required for both polarity and suppressiveness; (2) the genetic characterization of classes of petite mutants; followed by (3) the use of restriction endonculease analysis of mtDNA to detect and map the hypothetical duplications or deletions; (4) the isolation and mapping of new mutations in the polarity region of mtDNA; (5) the analysis of mutant alleles of the omega locus. We propose that yeast mitochondria are a good system for investigating the general phenomenon of recombination involving duplications, deletions and frameshift mutations. Additionally these studies will contribute new insights to the fine structure mapping of yeast mtDNA.